<![CDATA[Late-industrializing states routinely pursue defence industrial autonomy through procurement-linked technology transfer, yet the resulting capability gains are persistently disappointing. The dominant policy response has been to tighten offset and local content obligations. This article argues that such a response misdiagnoses the constraint. Drawing on absorptive capacity theory, technological capability accumulation, and the economics of military-technology diffusion, it develops the Defence Technology Integrative Capability (DTIC) framework, which specifies defence-industrial capability as the integration of four complementary assets—technoware, humanware, orgaware, and infoware—and argues that their returns are multiplicative rather than additive. The framework's central claim is that capability accumulation is governed by the least-developed asset rather than the average, so that expenditure directed at the wrong margin yields negligible returns. The argument is illustrated with Indonesia, a case selected because it inverts the conventional narrative: SIPRI and World Bank data show Indonesia sustains an expansive defence-industrial programme on a military burden of 0.78 per cent of GDP, gross domestic expenditure on R&D of 0.28 per cent of GDP, and 395 researchers per million inhabitants—roughly one twenty-fourth of the Republic of Korea's research density. No Indonesian firm appears in the SIPRI Top 100. A structured focused comparison with Japan, the Republic of Korea, Türkiye, and India indicates that in each successful case, investment in absorptive infrastructure preceded rather than followed industrial expansion. Seven propositions are derived, an operational measurement instrument is specified, and the implications for offset policies in low-absorptive-capacity economies are drawn. The analysis suggests that offset stringency and local content mandates are weak instruments where the human and organizational base cannot convert transferred knowledge into design capability, and that sequencing—not spending—is the binding policy variable.]]>
